Hawaii Extended Forecast Discussion
NWS Weather Prediction Center College Park MD
725 AM EST Tue Nov 14 2017

Valid 00Z Wed Nov 15 2017 - 00Z Wed Nov 22 2017

For much of the forecast period, Hawai`i will be along the
southwestern flank of a longwave trough which will continue to
reload during the week. The pattern should remain rather
consistent as a mean upper ridge remains a fixture in the forecast
in the vicinity of the International Dateline. While models agree
on the timing/position of the initial longwave trough, they differ
with the details of the second system late in the work week. By
Friday, ensemble spaghetti plots show it is essentially the 00Z
ECMWF/ECMWF ensemble mean versus all other models. By Saturday
morning, most of the 00Z ECMWF ensemble suite show a closed low to
the north of the island chain while other solutions favor a more
eastern scenario. The former camp would favor a retrograding upper
low which is in stark contrast to other models. The ECMWF has
shown consistency in such a forecast although the 00Z run is east
of its preceding cycle. While the ensemble means do differ, it is
not as extreme as the operational ECMWF. Would suggest an ensemble
means based approach taking a split between the 06Z GEFS/00Z ECMWF
ensemble means by as early as Friday as guidance quickly diverge.

At the surface, a prominent ridge across the central Pacific will
keep a rather tight pressure gradient in place. A weakening
boundary to the south of the Hawai`ian islands will maintain a
east-northeasterly flow regime. Forecast winds will be in the 20
to 25 knot range at times early on with a brief slackening
possible before the next surface high builds to the north. By
Thursday afternoon, some of the guidance show a 1044-mb ridge
setting up south of the Aleutians which will gradually help
intensify the local wind fields again. By Friday onward, model
differences become challenging as the 00Z ECMWF would suggest a
convergence zone associated with a front sweeping through the
region. Thus, it could either be weakening east-northeasterly flow
as depicted by most models or variable winds in response to the
nearby surface boundary. Regarding precipitation, the general
east-northeasterly winds will bring scattered showers through the
island chain with an eventual decrease in coverage as precipitable
water values drop by mid-week. If the ECMWF verifies, it could
become quite wet late in the period as moisture increases ahead of
the meandering upper low. However, other solutions would favor
scattered showers, thus supporting the general dichotomy here.


Rubin-Oster